2023-01-19: APP 2.0.0-beta13 has been released !
!!! Big performance increase due to optimizations in integration !!!
and upgraded development platform to GraalVM 22.3 based on openJDK19
We are very close now to releasing APP 2.0.0 stable with a complete printable manual...
Improvements for issues with wide field milky way panoramas
I do a lot of tracked milky way wide field panoramas (e.g. 3x3 mosaic with a 50mm or 85mm lens on a full frame camera) which I later combine with separate foreground shots and a lot of the sky shots have some parts of the foreground (horizon line, tree branches, ...) in it as I usually capture the sky from the same position as the foreground.
As all of the panorama stitching software out there I know of (e.g. Hugin, PTGui, ...) have some serious problems in stitching night sky images which are hard to correct manually, especially when it comes to seamless transitions between the frames when removing the stars and stretching the image (hard edges along the seam lines in the starless part or edges of different star shapes etc.) I really like to use APP for that.
APP does a really good job at this, far better than other stitching software, but this seems only to be the case when the field of view is not too big and there are not too much obstacles in the images. With a 85mm 3x3 shot and only some little foreground in the lower row it works really well but e.g. in a 3x3 50mm shot with a lot of tree branches around the result I get from APP is not useable at all.
What I wish I could do here is apply some masking on the source images to tell APP that there is no sky in the appropriate parts which really could the result here.
What also could be nice is to have some options for the panoramic image projection (rectlinear, which seems to be used in APP, or alternatively mercator, spheric, cylindric, etc.).
I know that APP is mainly designed for deep sky imaging, but finding good stitching software for wide field is really hard and APP seems to have really good potential here.
Trees and such can cause an issue indeed, we do want to improve on this and is on our to-do list. Not with a very high priority at the moment though.